Artificial Analysis Coding Agent Index v1.1
Progress Over Time
Interactive timeline showing model performance evolution on Artificial Analysis Coding Agent Index v1.1
Artificial Analysis Coding Agent Index v1.1 Leaderboard
| Context | Cost | License | ||||
|---|---|---|---|---|---|---|
| 1 |  GPT-5.6 SolNew OpenAI | — | 1.1M | $5.00 / $30.00 | ||
| 2 | OpenAI | — | 1.1M | $2.50 / $15.00 | ||
| 3 |  Grok 4.5New xAI | — | 500K | $2.00 / $6.00 | ||
| 4 | GPT-5.6 LunaNew OpenAI | — | 1.1M | $1.00 / $6.00 |
What is Artificial Analysis Coding Agent Index v1.1?
The Artificial Analysis Coding Agent Index is an independent composite score for coding-agent performance that pairs models with agentic harnesses. Version 1.1 is a simple average of pass@1 across three components: DeepSWE, Terminal-Bench v2, and SWE-Atlas-QnA.
Artificial Analysis Coding Agent Index v1.1 is a text benchmark evaluating models on agents and code tasks. LLM Stats tracks 4 models on this benchmark, scored on a 0–100 scale. The current average is 77.0, with the leader at 80.0.
Compare leaders on the best AI for agents and best AI for code leaderboards.
Current leaders
GPT-5.6 Sol from OpenAI currently leads the Artificial Analysis Coding Agent Index v1.1 leaderboard with a score of 80.000 across 4 evaluated AI models.
FAQ
Common questions about the Artificial Analysis Coding Agent Index v1.1 benchmark and leaderboard.